1. Field of the Invention
This invention relates in general to light measuring, and, more particularly, to means for providing a continuous real-time indication of the radiated energy output of a laser by accurately sampling and measuring beam power of a laser without blocking or unduly perturbing the laser beam.
2. Description of the Related Art
Measurement of the power of the light beam emitted by commonly available laboratory lasers, especially of helium-neon lasers, is a routine task that is usually performed by laboratory personnel to what is believed to be a high level of accuracy.
This belief, as to the level of accuracy of the reported measurements, is quite often erroneous and sometimes incorrect up to thirty percent (30%) due to the measurement means currently available.
Approximately twelve years ago, as helium-neon lasers became more readily available for laboratory use, the National Bureau of Standards and the American Society for Testing and Materials conducted intercomparisons of the output power measurements being made by eleven laboratories on the output of helium-neon lasers. This comparison study revealed differences in the reported measurements of beam power of plus or minus 18 percent to be typical, and plus or minus 59 percent to be the extreme.
As more experience was gained both by measurement personnel and test instrument designers, the uncertainty margin for error in reported measurements began to decline.
Thermal detectors, which had long been used for measuring high power levels and uniform Infrared radiation from lamps, had to be re-designed to measure a concentrated, lower powered laser beam. Silicon photodiodes used in the thermal detectors however, had in some cases too much power output and a non-uniform response curve over the spectral distribution range of the laser beam's power envelope. Later testing and experimentation proved that the reason for many of the shortcomings of the silicon photodiodes was due to the fact that the very coherence of the laser beam radiation was causing an interference pattern to be set up in the flat glass protective windows encasing the silicon photodiodes, thus varying the beam's reflectance by upwards of seven percent.
There were also measurement problems caused by adding a metal film neutral density filter to attenuate the beam, or an interference filter to eliminate ambient light. Both of these elements require further measurement corrections to obtain a valid measurement of beam power, as they add more flat glass windows, polarization sensitivity and inter-reflection of the beam to the original system. Also, the additional complexity of the measurement path led to critical path alignment problems which, if uncorrected, would give erroneous results.
Many of these problems are lessened by adding translucent diffusers, that is, discs of flashed opal glass or white plastic, to the measurement apparatus. These translucent diffusers both attenuate the radiation reaching the detector element and destroy the coherence of the beam, so that reflectance non-uniformity of the beam at the point of measurement is no longer considered to be a problem.
Other forms of diffuser elements are also found in the prior art, whereby the radiation is diffusely reflected throughout an integrating sphere, that is, a sphere painted with a highly reflective, diffuse white paint. A small sampling port is provided in the sphere for the detector element and is positioned so as to align the detector element to avoid direct primary beam reflection.
Measurement uncertainty using these designs eventually became limited by the standard detector element used to calibrate the design.
In 1984, researchers at the National Bureau of Standards investigating the characteristics of silicon photodiodes, discovered that a particular type of silicon photodiode was internally 100 percent quantum efficient, that is, it had no losses in converting light to photocurrent over the entire visible wavelength region, other than that light lost due to surface reflection. By using four photodiodes aligned to inter-reflect an incident beam and summing the detector currents, a power level measurement within plus or minus 0.1 percent of the theoretical absolute value was attained.
Although it would appear that the problems of beam power measurement of laboratory lasers might disappear, they did not. The inherent characteristics of lasers themselves remain a problem in the areas of power instability and, in an even more subtle problem area, polarization of the beam.
A gas filled laser, such as the helium-neon type, is available in two distinct versions, both of which are highly polarized.
One of the two types is called "randomly polarized", or sometimes by the gross misnomer "unpolarized", and is characterized by being polarized in two orthogonal axes. Beam power alternates from one polarization plane to the other in a repetitive pattern that varies with time. Orientation of the polarization axes is set during manufacture of the laser device not by plan, but rather due to the random asymmetries in the resonant cavity optics. As the laser tube approaches thermal equilibrium with its environment, the rate of polarization exchange between the two axes tends to slow with respect to time, but the polarization orientation of the beam at any one moment is indeed random.
The second version of the gas filled laser inserts a Brewster window in the resonant cavity of the laser. This Brewster window is usually a flat glass window which, when tilted at a particular angle, is largely reflective to one axis of polarization and completely transparent to the other. Thus, power on only one axis is encouraged to resonate, and the result can be a completely linearly polarized beam. Variations in the radiated power, however, can be plus or minus 7 percent or more, depending on design, and while the variations may be rapid when the laser is first activated, they will tend to slow with respect to time but never disappear, as the resonant cavity approaches thermal equilibrium.
As a laser beam passes through optical elements, such as the Brewster window, reflectance will vary with orientation of the polarization axis of the beam and will even be affected by dielectric coatings, lens angles, centering alignment, and even residual optical strains. Each and every one of these elements will create further variances in power.
With the above background to the problem in mind, it is seen why modern measuring devices are inadequate and inaccurate in measuring beam power.
In general, when measuring beam power with a radiometer, the beam is temporarily blocked. When this blockage is removed, changes in beam power may occur and go undetected. The present invention provides a beam power monitor that solves both of these problems, that is, providing a constant indication of beam power without interrupting the beam.
While beamsplitters are not uncommon as beam power monitors, they have previously been designed to either avoid polarization by reflecting the beam from an optical wedge at a small angle so as to be polarization insensitive, or by relecting the beam at the Brewster angle with the requirement that polarization be invariant.
An optical chopper can be modified to monitor beam power by making one side of the blade a mirror to reflect the beam back to a detector whenever the beam is blocked, but this has limited applications.
While all these existing methods and devices attempt to avoid being polarization sensitive, when they succeed, they do lose valuable information regarding polarization effects within the laser system itself.
Accordingly, the present invention addresses these problems presently found in laser beam power measuring devices and provides not only a solution, but a means that is unobtrusive and accurate over a wide-range of power levels.